Biochemistry 2005, 44, 16079-16089 16079

Heme Binding by the N-Terminal Fragment 1-44 of Human Growth Hormone†

Barbara Spolaore, Vincenzo De Filippis, and Angelo Fontana*
CRIBI Biotechnology Centre, UniVersity of Padua, Viale G. Colombo 3, 35121 Padua, Italy
ReceiVed July 15, 2005; ReVised Manuscript ReceiVed September 26, 2005

ABSTRACT: Fragment 1-44 of human growth hormone (hGH), prepared in vitro by limited proteolysis of
the hormone with pepsin at low pH, encompasses in full the N-terminal helix of this four-helix bundle
protein [Spolaore, B., Polverino de Laureto, P., Zambonin, M., and Fontana, A. (2004) Biochemistry 40,
9460-9468]. Here, we report the new and interesting observation that fragment 1-44 can bind heme.
The binding property is specific for the N-terminal helix of hGH, since heme binding does not occur with
fragment 45-191 or the entire protein. The spectral characteristics of Fe-protoporphyrin IX are those of
a low-spin, hexacoordinated iron ligated by two imidazole rings of His residues or His and Met residues.
See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

Far-UV circular dichroism (CD) measurements revealed that fragment 1-44 acquires a helical secondary
structure upon heme binding. Heme appears to be bound to the fragment in a stereospecific way, since an
induced dichroic signal is observed in the Soret region of the CD spectrum. The heme-fragment complex
occurs in a 1:1 molar ratio, as determined by spectrophotometric titration, as well as by electrospray-
ionization mass spectrometric analysis of the complex. The fragment alone is much more susceptible to
Downloaded via UNIV ICESI on August 13, 2020 at 21:00:51 (UTC).

tryptic digestion than the heme complex, implying a more folded and rigid structure of this last species.
It is proposed that the molecular features of fragment 1-44 determining its heme-binding property reside
in the amphipathic character of the helix adopted by the fragment, as well as in the presence in its
polypeptide chain of His18, His21, and Met14. These residues can act as specific ligands for the heme-
iron, as observed with cytochromes.

Heme (Fe-protoporphyrin IX) is a protein cofactor that
nature uses frequently to achieve different important func-
tions, such as reversible O2 binding, O2 activation, decom-
position of peroxides, and electron transfer (1). Moreover,
it has been shown that heme can regulate many biological
processes, including transcription, translation, protein trans-
location, and erythroid differentiation (2). Most of the natural
heme-binding proteins fold into R-helical structures and upon
removal of heme adopt a partly folded (3-6) or largely
unfolded (7, 8) state. The family of heme proteins has been
extensively studied, mostly with the aim to unravel the
protein structural features which are important for an efficient
binding of heme and related biological functions. Indeed,
heme binding imposes strong constraints to the structure of
a protein, which can be stronger than those caused by the
packing of a hydrophobic core (9). The major contribution
to the affinity of heme is due to a nonspecific partitioning
of heme into a hydrophobic region of a protein, which
accounts for the tendency of heme to adsorb to a variety of
proteins (10-15). Second, a specific binding of the heme is
dictated by the formation of coordination bonds with the
imidazole ring of histidine and/or other specific interactions
with amino acid residues surrounding the porphyrin ring (16,
17). Heme protein design is also nowadays intensively
pursued with the goal to better understand structure-function
relationships in heme proteins and to design molecules with
† MS, mass spectrometry; TFA, trifluoroacetic acid; Tris, tris(hydroxy-
This work was supported in part by the Italian Ministry of
University and Research (FIRB Project on Protein Folding and PRIN-
2003).
* To whom correspondence should be addressed. Tel.: +39-049-
8276156. Fax: +39-049-8276159. E-mail: angelo.fontana@unipd.it.
new desired functions (17-21). Protein helical scaffolds have
been mostly used by protein designers, due to the prevalence
of R-helical heme proteins in nature and the relatively easy
de novo design of helical polypeptides (22-28). However,
heme affinity measurements have demonstrated that even the
best designed protein moieties bind heme much more weakly
than natural heme proteins (17, 21).
Human growth hormone (hGH1) is a four-helix bundle
protein of 191 amino acid residues (29) that principally
regulates the body growth, but affects also the metabolism
of proteins, carbohydrate, and lipids (30-32). A proposal
was advanced that hGH acts also as a pro-hormone and,
therefore, that it requires its proteolytic cleavage to produce
various peptide fragments that display the diversity of
biological effects of hGH (33). It has been reported that two
fragments of hGH are present in vivo: fragment 1-43, which
shows insulin-potentiating effects, and fragment 44-191,
which has diabetogenic activity (34-36). These fragments
have been detected in serum and at the level of the pituitary
gland (34, 37, 38), but their mechanisms of formation from
the parent protein are not known. Since they are two
complementary parts of the 191-residue chain of hGH, it is
1
Abbreviations: hGH, human growth hormone; CD, circular dichro-
ism; ESI, electrospray ionization; E/S, enzyme-to-substrate ratio; ESI,
electrospray-ionization; Gdn‚HCl, guanidine hydrochloride; HPLC, high
performance liquid chromatography; kDa, kilodalton; RP, reverse-phase;
methyl)aminomethane; UV, ultraviolet; 1-44, N-terminal fragment of
hGH.; 45-191, C-terminal fragment of hGH cross-linked by two
disulfide bridges (Cys53-Cys165 and Cys182-Cys189); RR, resonance
Raman.

10.1021/bi051374d CCC: $30.25 © 2005 American Chemical Society

Published on Web 11/16/2005
16080 Biochemistry, Vol. 44, No. 49, 2005
conceivable to suggest that they are formed by selective
proteolytic cleavage of the native protein. However, at
present the specific protease responsible for their generation
has not been identified.
In a recent paper, we have shown that limited proteolysis
of hGH in vitro by pepsin at pH 4.0 leads to the production
of fragments 1-44 and 45-191, as a result of the selective
cleavage of the Phe44-Leu45 peptide bond (39). Therefore,
the two peptic fragments of hGH prepared in vitro are related
to the two physiologically relevant fragments occurring in
vivo. It has been found that fragment 45-191, encompassing
helixes 2, 3, and 4 of hGH and containing the two disulfide
bridges of the protein (29), adopts a helical conformation in
solution at neutral pH. Conversely, fragment 1-44, compris-
ing the first helix of the native hormone, is rather disordered.
Here, we report that the hGH fragment 1-44 of hGH can
bind heme with high specificity. This binding phenomenon
has been discovered by serendipity. Indeed, during a
purification step of the fragment by size-exclusion chroma-
tography, we have found that it was eluted complexed to
heme from a gel filtration column that had been inadvertently
polluted by heme in previous separations of myoglobin using
the same column. Upon binding of the heme moiety, the
fragment acquires a helical secondary structure, as judged
from far-UV circular dichrosim (CD) measurements. The
conformational properties of the heme-fragment complex,
as well as the spectroscopic features of the heme when bound
to the fragment, have been investigated. The molecular
features of the hGH fragment 1-44 that likely dictate its
heme-binding properties are discussed.
MATERIALS AND METHODS
Materials. Recombinant human growth hormone (hGH)
was produced in Bacillus subtilis (40) and stored as a
lyophilized sample at 4 °C. The protein was shown to be
homogeneous by reverse-phase high-performance liquid
chromatography (RP-HPLC) and by sodium dodecyl sulfate
(SDS) polyacrylamide gel electrophoresis (PAGE) and to
possess the correct N-terminal sequence. Pepsin, trypsin,
hemin, and sodium dithionite were obtained from Sigma.
Acetonitrile and trifluoroacetic acid (TFA) were purchased
from Fluka. All other chemicals were analytical grade and
were obtained from Sigma or Fluka. Water used to prepare
buffers was purified by a Millipore Milli-Q Plus system.
Preparation of Fragment 1-44 and Fragment 45-191.
Fragments 1-44 and 45-191 were produced by limited
proteolysis of hGH at pH 4.0 following essentially the
procedure described previously (39). Briefly, the reaction was
performed on a solution of hGH (0.25 mg/mL) in 10 mM
Na citrate/0.15 M NaCl, pH 4.0, at 4 °C with a protease:
substrate (E:S) ratio of 1:10 (by mass) for 2 h and then
stopped by alkalinization of the reaction mixture with 5%
aqueous ammonia. Purification of fragments 1-44 and 45-
191 from the limited proteolysis reaction mixture was
performed by RP-HPLC utilizing a Vydac C18 column (10
× 250 mm) purchased from The Separations Group (Hes-
peria, CA). The column was eluted at a flow rate of 2.5 mL/
min with a linear gradient of water and acetonitrile containing
0.1% and 0.085% TFA respectively, from 30% of acetonitrile
to 55% in 8 min and from 55% to 63% in 26 min. The
effluent from the column was monitored by measuring the
absorbance at 226 nm.
Spolaore et al.
All experiments were performed on fragment 1-44 and
fragment 45-191 recovered after gel filtration chromatog-
raphy. To this aim, both hGH fragments purified by RP-
HPLC were lyophilized using the Speed-Vac system (Sa-
vant), redissolved in 6 M Gdn‚HCl and loaded (250 µL, ∼1
mg/mL) on a Superdex-75 column (type HR 10/30, 1 × 30
cm; Pharmacia) equilibrated and eluted with 20 mM Tris‚
HCl/0.15 M NaCl, pH 7.5, at a flow rate of 0.4 mL/min.
The absorbance of the effluent was recorded at 226 nm. The
column was calibrated using a protein mixture kit of low
molecular mass for gel filtration chromatography (Pharma-
cia).
UV-Visible Absorption Spectroscopy. Protein and heme
concentrations were determined using a Perkin-Elmer Lambda-
25 spectrophotometer. The concentrations of hGH and
fragments 1-44 and 45-191 were determined from absor-
bance measurements at 280 nm according to Gill and von
Hippel (41). Stock heme solutions were prepared in 0.1 M
NaOH, and the precise concentration of heme was deter-
mined using an extinction coefficient for heme of 5.84 ×
104 cm-1 M-1 at 385 nm (42). Heme stock solutions were
then prepared in 20 mM Tris‚HCl/0.15 M NaCl, pH 7.5,
immediately before use. Reduction of the heme bound to
fragment 1-44 was conducted by adding to the solution of
the heme-fragment complex a 1000-fold molar excess of
sodium dithionite, taken from a freshly prepared stock
solution of the reagent.
Spectrophotometric titration of the binding of heme to
fragment 1-44 was performed as described by Tsutsui and
Mueller (43). A heme stock solution was added simulta-
neously into two cuvettes, one containing a solution 5.3 µM
fragment 1-44 in 0.1 M Tris‚HCl/0.15 M NaCl buffer, pH
7.5, and the other containing buffer only (reference cuvette).
The stock solution (0.58 mM) of heme in 0.1 M NaOH was
added at increments of 5, 10, 20, and 25 µL. After each
addition of heme, the samples in the two cuvettes were stirred
for 5 min and allowed to stand for an additional 5 min. Then,
difference absorption spectra were recorded. These spectra
had a maximum at 414 nm and a minimum at 368 nm. The
heme-binding curve was constructed by plotting the ∆A414 nm
versus the heme concentration. The titration of fragment
1-44 with heme was followed also by acquiring far-UV
circular dichroism (CD) spectra of the fragment solution after
successive additions of aliquots of a stock solution of heme.
A titration curve was derived by plotting heme concentrations
versus the variation of the mean residue ellipticity at 222
nm. The two heme-binding curves obtained by UV and far-
UV CD measurements were fitted to the equation for the
dissociation constant Kd assuming a 1:1 binding (44, 45).
For the curve fitting the Microcal Origin software (Microcal
Software Inc., USA) was used.
Circular Dichroism. CD measurements were made at 25
°C on a Jasco J-710 spectropolarimeter (Tokyo, Japan)
equipped with a thermostatically controlled cell holder. The
instrument was calibrated with d-(+)10-camphorsulfonic
acid. Far-UV CD spectra were recorded at a fragment
concentration of 0.028 mg/mL using a quartz cell with a path
length of 1 cm, and the results were expressed as mean
residue ellipticity [θ] (deg‚cm2‚dmol-1). Far-UV CD spectra
were analyzed in order to estimate the percentage of protein
secondary structure using the equation reported by Scholtz
et al. (46). The CD measurements in the 350-500 nm region
Heme Binding by the hGH Fragment 1-44
were obtained with a solution of fragment 1-44 (19.2 µM)
containing heme (38 µM) using a quartz cell of 1.0 cm path
length. The ellipticity (θobs) was expressed in millidegrees.
Resonance Raman Spectroscopy. Resonance Raman (RR)
spectra of the heme-fragment complex were recorded at
room temperature in 20 mM Tris‚HCl/0.15 M NaCl buffer,
pH 7.5, following essentially procedures before described
(47).
Electrospray Mass Spectrometry. Mass spectra were
acquired on a tandem mass spectrometer Q-TOF Micro
(Micromass, U.K.) equipped with a Z-spray nanoflow
electrospray-ionization (ESI) interface. Mass spectra (positive
ion mode) of tryptic peptides of fragment 1-44 were
acquired using the electrospray source operating at capillary,
cone, and extractor voltages of 3200, 45, and 1 V, respec-
tively. The temperature of the source was 80 °C and that of
the desolvation gas 250 °C. For the MS analysis, samples
were dissolved in a 1:1 solution of acetonitrile:water contain-
ing 1% formic acid.
The heme-fragment complex was analyzed by nano-
electrospray mass spectrometry (MS). Nano-ESI capillaries
were prepared in house from borosilicate glass tubes of 1
mm OD and 0.78 mm ID (Harvard Apparatus, Holliston,
MA) using a micropipet puller (Sutter Instruments, Hercules,
CA, model P-80 PC) and gold coated using an Edwards
S-150B sputter coater (Edwards High Vacuum, West Sussex,
U.K.). Typically, 0.5-2 µL of solution were analyzed. For
the MS analysis the following experimental parameters were
used (positive ion mode): capillary voltage, 1.7 kV; sample
cone, 30.0 V; extractor cone, 1.0 V. In the collision cell argon
was at an inlet pressure of 10 psi and the collision energy
setting was 4.0 V. The ESI source temperature was set at
30 °C.
MS/MS analyses of the heme-fragment complex were
performed utilizing the same parameters of the MS instru-
ment as above. The collision gas was argon at an inlet
pressure of 10 psi, and the collision energy setting was 45
V. Instrument control and data acquisition and processing
were achieved using the MassLynx software (Micromass,
U.K.).
MS analyses on fragment 1-44 were made utilizing a
fragment sample which was recovered after gel filtration
chromatography using a Superdex-75 column equilibrated
and eluted with 0.1 M ammonium acetate, pH 7.5. Samples
of fragment 1-44 in acetate buffer (10.6 µM) were mixed
with aliquots of a freshly prepared solution of heme in 0.01
M NaOH (88 µM). Mass spectra of all fragment samples
were acquired after 30 min incubation at room temperature.
Limited Proteolysis Experiments. Fragment 1-44 in the
presence or absence of heme was subjected to limited
proteolysis with trypsin. The reactions were performed in
parallel on two solutions of fragment 1-44 (0.07 mg/mL)
in 20 mM Tris‚HCl/0.15 M NaCl, pH 7.5, one in the presence
of heme at a molar ratio 1:3 of fragment over heme and the
other in the absence of heme. After 30 min incubation,
trypsin was added to both solutions at a protease:substrate
(E:S) ratio of 1:500 (by weight). Aliquots were taken from
the two reaction mixtures at intervals, and then proteolysis
was quenched by acidification with an aqueous solution of
5% TFA. Samples were loaded onto a Zorbax Eclipse XDB-
C8 column (4.6 × 150 mm; Agilent Technologies, Palo Alto,
CA) equilibrated with a mixture of 20% acetonitrile in water
Biochemistry, Vol. 44, No. 49, 2005 16081
FIGURE 1: RP-HPLC analysis of the proteolytic mixture of hGH
with pepsin. Proteolysis was conducted for 2 h at 4 °C (E/S 1/10,
by weight) in 10 mM Na citrate/0.15 M NaCl buffer, pH 4.0. The
chromatographic peaks containing hGH and fragments 1-44 and
45-191 are indicated in the chromatogram. A Vydac C18 column
(10 × 250 mm) was employed as described under Materials and
Methods.
containing 0.085% TFA and eluted with a linear gradient of
acetonitrile from 20% to 52% in 22 min at a flow rate of
0.8 mL/min. The absorbance of the effluent was recorded at
226 nm. The percent recoveries during proteolysis of
fragment 1-44 and fragment 1-38 (which is formed in trace
amounts and elutes at the same retention time of the intact
fragment) and fragment 1-41 were determined by integrating
the area under the HPLC peak of each fragment.
RESULTS
Limited proteolysis of hGH with pepsin in 10 mM Na
citrate/0.15 M NaCl, pH 4.0, at 4 °C occurs selectively at
the level of the peptide bond Phe44-Leu45 (39). The
reaction is quite clean, and it allows the selective production
of fragment 1-44 and fragment 45-191, as shown by the
RP-HPLC chromatogram of an aliquot of the reaction
mixture (Figure 1). It was convenient to stop the proteolysis
reaction after 2 h, in order to minimize the subsequent
proteolytic degradation of the two fragments, especially
fragment 1-44. Both fragments were purified first by RP-
HPLC and then by gel filtration chromatography utilizing a
Superdex-75 column equilibrated and eluted with 20 mM
Tris‚HCl/0.15 M NaCl buffer, pH 7.5 (not shown). The gel
filtration step was used for separating the fragments from
some aggregated material, as well as for preparing suitable
solutions of homogeneous fragments for subsequent studies.
The homogeneity of the fragments thus produced was
established by a variety of analytical criteria, including
electrospray-ionization (ESI) mass spectrometry (MS) (see
also ref 39).
The first indication that the hGH fragment 1-44 was able
to bind heme was derived from the fact that the fragment
solution prepared after gel filtration was reddish. Then, it
was realized that the colored solution was due to the fact
that the Superdex-75 column was earlier used by us with
samples of the heme-containing protein myoglobin. The
16082 Biochemistry, Vol. 44, No. 49, 2005
FIGURE 2: Absorption spectra of heme in the absence and presence
of fragment 1-44 (A), hGH (B), and fragment 45-191 (C). (A)
UV-vis spectra of a 4.5 µM solution of heme alone or in the
presence of fragment 1-44 (10 µM) at a 1:2.25 molar ratio of heme/
fragment. (B) UV-vis spectra of heme in the presence or absence
of hGH and (C) with or without fragment 45-191. The concentra-
tion of hGH and fragment 45-191 was 5.1 µM and 1.9 µM,
respectively. Both spectra B and C were recorded at a 1:1 molar
ratio of heme:protein. Measurements were taken at room temper-
ature in 20 mM Tris‚HCl/0.15 M NaCl, pH 7.5.
conclusion was that fragment 1-44 and heme, despite their
largely different molecular masses, eluted together from the
gel filtration column, likely as a heme-fragment complex.
The absorption spectrum of the complex eluted from the gel
filtration column showed typical spectroscopic features of
hemo-proteins (see below).
Absorption Spectra of Heme Bound to Fragment 1-44.
The heme-binding capacity of fragment 1-44 was confirmed
by adding a stock solution of Fe-protoporphyrin IX in 0.1
M NaOH to a solution of the fragment in Tris buffer, pH
7.5. The absorption spectrum of this solution was character-
ized by a Soret maximum at 413 nm and a broad band around
532 nm (Figure 2A). We investigated also the possibility
that the complementary fragment 45-191 or hGH could bind
the heme. To this aim, heme was added to solutions of these
two protein species dissolved in Tris buffer, pH 7.5, under
the same conditions used for fragment 1-44 and the
absorption spectra of the solutions were recorded. As shown
in Figure 2B,C, the features of the spectrum of heme do not
change in the presence of added hGH or fragment 45-191,
indicating the absence of a specific incorporation of heme
into these protein species. Therefore, the property of heme
binding is specific for fragment 1-44.
The absorption spectrum of heme bound to fragment 1-44
shows significant changes when reduced with sodium
dithionite. As shown in Figure 3, the Soret band becomes
slightly sharper and red-shifted to 425 nm, whereas in the
R/β band region there is a pronounced splitting resulting in
a well-resolved β-band at 530 nm and an R-band at 560 nm.
These values are typical of bis-histidine ligated b-type
cytochromes or of heme coordinated by one histidine and
one methionine (20). Moreover, these spectral features are
those of iron in a low-spin and hexacoordinated state (24,
48) (see also Discussion).
Heme-Fragment Dissociation Constant. The fact that
fragment 1-44 retains a bound heme even after a gel
filtration chromatographic step is a clear-cut indication of a
strong affinity of heme for the fragment. Indeed, we
measured the dissociation constant Kd between fragment
Spolaore et al.
FIGURE 3: Absorption spectra of oxidized (- - -) and reduced
(s) heme complex of fragment 1-44. UV-vis spectra of fragment
1-44 with oxidized and with reduced heme at a 0.6:1.0 molar ratio
of heme:fragment (excess of fragment). Spectra were recorded in
20 mM Tris‚HCl/0.15 M NaCl buffer, pH 7.5 at room temperature
and a fragment concentration of 7.9 µM. Heme bound to fragment
1-44 was reduced by addition of a 1000-fold molar excess of
sodium dithionite. The maxima in the two spectra are indicated.
FIGURE 4: Titration of fragment 1-44 with heme. Difference
absorption spectra acquired at increasing concentrations of heme
over fragment 1-44. A hemin solution (0.58 mM in 0.1 M NaOH)
was simultaneously added to a 5.3 µM solution of fragment 1-44
in 0.1 M Tris‚HCl/0.15 M NaCl buffer, pH 7.5, and a reference
cuvette containing Tris buffer only. After each addition of the heme
solution, both the reference and the sample were stirred for 5 min
and then allowed to stand for 5 min before spectra were acquired.
The titration was performed at room temperature. The inset shows
the heme-binding curve generated by plotting the difference
absorbance at 414 nm versus the heme concentration.
1-44 and heme, assuming that in solution heme does not
exist in an aggregated state, but only in the free or bound
state. A solution of fragment 1-44 at a fixed concentration
was titrated with increasing amounts of heme, and difference
absorption spectra were recorded against a solution of just
buffer and heme. The increase in absorbance at 414 nm
versus the heme concentration (Figure 4) was fitted to an
equation for a 1:1 binding (44), and a figure of 1.48 µM (
0.13 was calculated for the Kd of the complex.
Circular Dichroism Measurements. An interesting feature
of the heme-fragment complex is that there is a change in
the peptide secondary structure of fragment 1-44 upon heme
binding. The far-UV CD spectra of the fragment with or
without heme are shown in Figure 5. The spectrum of the
complex has been acquired in the presence of an excess of
heme (two equivalents), in order to shift the heme-fragment
equilibrium toward the bound form. Whereas the CD
Heme Binding by the hGH Fragment 1-44
FIGURE 5: Far-UV CD spectra of fragment 1-44 in the absence
and presence of heme. The molar ratio between fragment 1-44
and heme was 1:1.8 (excess of heme) at a fragment concentration
of 5.3 µM. Spectra were recorded at 25 °C in 0.1 M Tris‚HCl/0.15
M NaCl buffer, pH 7.5. The inset shows the titration curve obtained
by plotting the ∆[θ]222 nm measured after each addition of heme to
a solution of fragment 1-44 (5.3 µM) versus the heme concentra-
tion. The curve fitting was performed as described under Materials
and Methods.
spectrum of fragment 1-44 in the absence of the Fe-
protoporphyrin IX is that of a largely unfolded polypeptide,
in the presence of heme the spectrum is characterized by
minima of ellipticity at 208 and 222 nm, which are indicative
of a significant degree of helicity (49). The far-UV CD
spectra were analyzed in terms of percent helical secondary
structure, and 9% and 26% helical content was calculated
for free fragment and heme-fragment complex, respectively
(46). However, it is known that estimates of percent helicity
in polypeptides with a low content of secondary structure,
as in the present case of fragment 1-44 alone, are not reliable
(49). Moreover, the 26% of the heme-fragment complex
can be in error, since the contribution of aromatic amino
acid residues and chromophores to the far-UV CD spectra
of polypeptides has been well documented, including that
of the heme chromophore (50, 51). In particular, it has been
concluded in these previous studies that the heme contribu-
tion to the far-UV CD spectrum of a heme-polypeptide leads
to an underestimate of the actual helical content of the
peptide/protein species.
The induction of secondary structure in fragment 1-44
upon heme binding was followed also by measuring the mean
residue ellipticity at 222 nm at increasing concentrations of
Fe-protoporphyrin IX. The data of ellipticity at 222 nm of
the fragment versus the concentration of heme (Figure 5,
inset) were fitted to the equation reported by Kuwabara et
al. (44, see also ref 45), and a Kd value of 2.33 µM ( 0.31
was estimated. This figure is quite similar to that obtained
by absorption measurements (Figure 4). It may be that the
slight difference can be attributed to the contribution of heme
to far-UV CD ellipticities (see above), that can affect the
estimate of Kd.
Symmetric chromophores that do not absorb circularly
polarized light become optically active when they are
embedded in a specific stereochemistry within a protein fold
(52). This acquired optical activity is defined as induced
circular dichroism (ICD), and it allows interesting protein-
Biochemistry, Vol. 44, No. 49, 2005 16083
FIGURE 6: Induced CD spectrum in the Soret region of heme
complexed with fragment 1-44. Spectra were recorded at 25 °C
in 20 mM Tris‚HCl/0.15 M NaCl buffer, pH 7.5. The concentration
of heme was 38 µM, and the molar ratio of fragment:heme was
1:2 (excess of heme). The fragment concentration was 19.2 µM.
ligand binding studies (53). Heme does not have a signal in
the visible region of the CD spectrum, if this moiety is free
in solution or associated at the surface of a protein (54).
However, upon binding to a protein in a chiral and relatively
rigid environment, heme gives rise to characteristic ICD
spectra. The optical activity of heme bound to fragment 1-44
was investigated in the Soret region (Figure 6). The CD
spectrum is characterized by a maximum at 400 nm and a
minimum at 420 nm, indicating that heme is bound to the
fragment inside a relatively rigid pocket (52).
Resonance Raman Spectroscopy. In an attempt to possibly
determine the amino acid side-chain residues of fragment
1-44 that act as ligands for the heme iron, resonance Raman
(RR) measurements on the heme-fragment complex were
carried out using an excitation wavelength in the Soret band
at 413.1 nm (47). The high (ν4, ν3, ν2, ν10 at 1373, 1504,
1576, and 1639 cm-1, respectively) and low (ν9 and ν8 at
268 and 346 cm-1) frequency region spectra of the complex
showed frequencies typical of an Fe(III) in a hexacoordinated
low spin state (not shown) (see ref 25). However, a careful
analysis of the RR spectra did not enable us to distinguish
between a bis-His or a Met/His axial heme coordination (see
Discussion).
Mass Spectrometry of the Heme-Fragment Complex. The
stoichiometry of the binding between heme and fragment
1-44 was investigated also by electrospray ionization (ESI)
mass spectrometry (MS). Indeed, recent advances of this
technique and, in particular, the use of nano-electrospray
allow one to analyze noncovalent protein complexes in
solution at physiological pH and to measure their masses
(55, 56). Noncovalent interactions can be kept intact during
the ionization process by carefully adjusting the parameters
(mainly voltages and temperatures) of the mass spectrometer
ion source.
For the ESI-MS analysis, we prepared a solution of heme
and fragment 1-44 in 0.1 M ammonium acetate, pH 7.5, at
a molar ratio of 2.0 (excess of heme) and we confirmed the
formation of the complex under these solvent conditions by
acquiring a UV-visible spectrum, which showed the typical
16084 Biochemistry, Vol. 44, No. 49, 2005 Spolaore et al.

FIGURE 7: Mass spectrometric analysis of the heme-fragment 1-44 complex. (A) Electrospray mass spectrum of the heme-fragment
complex. The identities and the charge states of each peak are indicated. The inset shows the deconvoluted mass spectra of (A) and the
calculated and measured molecular masses of the various components. (B) MS/MS analysis of the charge state +4 of the heme-fragment
complex. The identities and charge states of the dissociated species are indicated. The complex was analyzed in 0.1 M ammonium acetate
buffer, pH 7.5, and the molar ratio heme:fragment was of 2:1 (excess of heme). The fragment concentration was 10 µM.

maxima of a hexacoordinated heme (see Figure 3 and
Discussion). Analysis of this solution by nano-electrospray
MS revealed the presence of multi-charge state peaks of
fragment 1-44, together with a species of higher mass
(Figure 7A). In particular, the measured mass of this last
species, 5978.8 Da (Figure 7A, inset), corresponds to the
calculated mass of the complex formed by fragment 1-44
(5363.1 Da) and heme (616.1 Da) at a 1:1 stoichiometry
(calculated mass of the complex 5979.2 Da). In the mass
spectrum there are no other species arising from a different
stoichiometry of association between the fragment and the
Fe-protoporphyrin IX. Moreover, we tested the ability of MS
measurements to assess the specificity of heme binding by
fragment 1-44. To this aim, we analyzed by MS a solution
of hGH, fragment 45-191, and a tryptic digest of hGH in
the presence of heme and we verified that heme binding does
not occur (data not shown). On the other hand, MS and MS/
MS analyses of a solution of apomyoglobin and heme gave
clear-cut data of heme binding, as expected from the well-
known binding of heme to apomyoglobin in a 1:1 molar ratio
(13) (not shown).
In order to confirm the identity of the heme-fragment
complex, we analyzed by MS/MS the charge state 4+ of
the species of 5978.8 Da, which is the ion at 1495.64 m/z
(see Figure 7A). Indeed, the MS/MS technique allows one
to select an ion of interest and, in the case of a noncovalent
complex, to dissociate it inside the mass spectrometer into
its components and to measure their masses. The MS/MS
spectrum thus acquired is shown in Figure 7B. It can be
observed that the ion at 1495.64 m/z dissociates into a species
at 616.1 m/z, which is the singly charged free ferric heme
ion [Fe(III)heme]+, and a species at 1788.27 m/z, that
corresponds to the charge state +3 of fragment 1-44 without
heme. Of interest, the ion +4 of the heme-fragment complex
loses the positive charge associated to heme, thus leading to
a triply protonated ion of the free fragment (57). Therefore,
MS analysis confirmed the identity of the species of 5978.8
Da as the noncovalent complex between fragment 1-44 and
heme at a 1:1 molar ratio.
Probing Fragment Structure by Proteolysis. We tested if
a proteolytic probe could monitor the enhancement of
secondary structure in fragment 1-44 upon heme binding,
Heme Binding by the hGH Fragment 1-44
FIGURE 8: Limited proteolysis of fragment 1-44 in the absence
or presence of heme with trypsin. (Top) Amino acid sequence of
fragment 1-44. The location of the helix is indicated by a gray
box (residues 9-34), whereas arrows indicate the sites of cleavage
by trypsin. (A) RP-HPLC analysis of the proteolysis mixtures of
fragment 1-44 with heme and without heme conducted at room
temperature (E/S 1/500, by weight) in 20 mM Tris‚HCl/0.15 M
NaCl buffer, pH 7.5, at a molar ratio of 3:1 heme:fragment 1-44
(excess of heme). The fragment concentration was 0.07 mg/mL.
Tryptic peptides separated by RP-HPLC were identified by ESI-
MS. (B) The percent recovery of fragment 1-44 (in the absence
of heme, open symbols)) and of fragments 1-44, 1-41, and 1-38
(in the presence of heme, closed symbols) during proteolysis
reported vs the reaction time of the proteolysis reaction (see
Materials and Methods).
as detected by far-UV CD measurements (see above Figure
5). Indeed, it is known that proteases easily attack unfolded
or partly folded proteins, while folded native proteins are
quite resistant to proteolysis (8, 58-60). Trypsin was chosen
as a proteolytic probe, considering that the fragment contains
up to five residues (Lys and Arg) which can be sites of tryptic
attack, evenly distributed along its 44-residue chain (Figure
8, top). Solutions of fragment 1-44 alone or in the presence
of heme at a 1:3 molar ratio of fragment:heme were digested
in parallel with trypsin, and the time course of the digestion
was determined by means of RP-HPLC analyses of aliquots
taken at intervals from the reaction mixtures. The indentity
of the various tryptic fragments was established by ESI-MS
(Table 1). As shown in Figure 8A, the fragment alone is
much more easily digested than the heme-fragment com-
plex. Of interest, in the absence of heme, the tryptic peptides
Biochemistry, Vol. 44, No. 49, 2005 16085
Table 1: Molecular Masses of Tryptic Peptides of Fragment 1-44a
tryptic peptides of molecular mass (Da)
fragment 1-44 observed calculated
1-44 5363.1 5363.1
1-41 4965.8 4965.7
1-38 4580.1 4580.2
9-44 4450.1 4451.0
17-44 3488.4 3487.7
20-44 3123.8 3123.5
17-41 3091.8 3090.5
20-41 2726.5 2726.3
20-38 2341.4 2341.1
1-16 1890.2 1890.0
9-16 978.6 978.5
1-8 929.6 929.5
42-44 415.2 415.2
a
Tryptic peptides of fragment 1-44 were obtained by limited
proteolysis of the fragment with trypsin at pH 7.5 in the presence or
absence of heme, as described in the text. The mass spectrometric
measurements were performed on the peptides isolated after RP-HPLC
of the proteolytic mixtures. Experimental molecular masses were
determined by ESI-MS, whereas theoretical average molecular masses
were calculated from the amino acid sequence of fragment 1-44.
produced from fragment 1-44 derive from cleavages at all
but one basic amino acid residue (Lys and Arg), since a
tryptic peptide deriving from the hydrolysis of the Lys38-
Glu39 peptide bond is not seen in the RP-HPLC chromato-
gram (Figure 8A). This derives from the fact that tryptic
cleavages are hindered at basic residues flanked by acidic
(Glu and Asp) residues (61). While the free fragment is fully
digested by trypsin after 15 min reaction, in the presence of
heme the fragment is almost fully resistant to proteolysis
and only a small amount of fragment 1-41 is formed.
Actually, MS analysis revealed that trace amounts of
fragment 1-38 are also formed, this last fragment species
being eluted together with the intact fragment from the RP-
HPLC column. After 3 h reaction, other tryptic fragments
are produced, but still fragment 1-44 remains quite abundant
in the proteolysis mixture. Therefore, the proteolytic probe
appears to trim the unstructured C-terminal end of the heme-
fragment complex, in agreement with the fact that the heme-
fragment complex adopts a hydrogen-bonded, helical sec-
ondary structure and that the first helix in intact hGH
encompasses the chain segment 9-34 (29). Of note, the
structural features of the N-terminal region of the fragment
cannot be probed by trypsin, since with this specific protease
peptide bond fission can occur only at Arg8 (see Figure 8,
top). Upon prolonged proteolysis, such as 3 h reaction, the
heme-fragment complex is slowly digested also at Arg16
and Arg19, both residues being located inside the helical
chain segment 9-34. It is reasonable to propose that the
equilibrium between the folded heme-bound fragment and
the unfolded unbound species actually controls the proteoly-
sis events, so that also the unfolded free fragment can act as
tryptic substrate. The different rates of proteolysis of
fragment 1-44 alone or in the presence of heme are better
highlighted by calculating the percent decrease of intact
fragment during proteolysis, as shown in Figure 8B. While
fragment 1-44 alone is completely digested in few minutes,
in the presence of heme the fragment is hydrolyzed much
more slowly, concomitantly producing mostly fragment
1-41. If we consider the total percent of fragments 1-44,
1-41, and 1-38, we can observe that their combined amount
16086 Biochemistry, Vol. 44, No. 49, 2005
decreases only about 50% after 5 h of proteolysis of the
fragment in the presence of heme (Figure 8, bottom).
DISCUSSION
In this paper, we report the novel observation that the hGH
fragment 1-44 can bind the heme moiety. The binding
property is not shared by fragment 45-191, nor by the entire
hGH molecule. The spectrophotometric titration of fragment
1-44 with heme allowed us to deduce as 1:1 the molar ratio
of the heme-fragment complex and a Kd of 1.48 µM.
Additional evidence for both complex formation and stoi-
chiometry was derived from ESI-MS measurements con-
ducted on the complex at neutral pH. Clear-cut MS evidence
of the 1:1 complex was obtained, since two main species
were detected in the MS spectrum with a difference of their
deconvoluted masses of 616 Da, corresponding to the mass
of heme. Moreover, the MS/MS analysis of the complex
resulted in the dissociation of the heme and concomitant
appearance of the MS peaks of fragment and heme. No
evidence for binding of two hemes to the fragment was
obtained by MS measurements. The rather tight binding of
heme to the fragment can be deduced, besides from the Kd
value of the complex, from the fact that the complex can be
separated by gel filtration chromatography. Moreover, the
induced CD in the Soret region of the heme moiety, which
is optically inactive in its free state in solution, is clearly
indicative of a stereospecific binding of the heme in a quite
rigid pocket of the folded fragment (52).
The interaction of heme with fragment 1-44 induces a
significant amount of R-helical secondary structure in the
otherwise largely unfolded fragment, as deduced from far-
UV CD measurements. Assuming that the helix adopted by
the fragment upon heme addition is similar to that of the
corresponding chain region in the four-helix bundle hormone,
the heme-mediated helix is expected to involve the chain
segment 9-34 (helix A) (29) and to be strongly amphiphilic,
as shown in Figure 9. It can be proposed that the hydrophobic
face of the amphiphilic helix, made up mostly by Ala, Phe,
and Leu residues, represents the docking site of the hydro-
phobic heme moiety (11, 62). As a matter of fact, in several
studies it has been demonstrated that an amphiphilic helix
is a key structural motif of synthetic polypeptides that are
capable of heme binding (20, 63). A number of peptide
models were constructed in such a way to form an am-
phiphilic helix that can bind heme. The hydrophilic and
hydrophobic face of the helix contained mostly polar and
charged (Glu, Lys) and hydrophobic (Phe, Leu) residues,
respectively (see ref 20 for a review).
Limited proteolysis experiments conducted on the heme-
fragment complex clearly demonstrated that the fragment
adopts a quite rigid structure not amenable to an easy
degradation by trypsin (8, 58-60). Proteolysis experiments
were conducted using trypsin as proteolytic probe, consider-
ing that this substrate-specific enzyme can cleave at the
C-terminus of Arg8, Arg16, Arg19, Lys38, and Lys41 (see
Figure 8, top). Residues 16 and 19 are roughly in the middle
of the first helix in native hGH (chain segment 9-34; 29)
and near the residues which can coordinate heme (Met14,
His18, and His21). The folded, heme-bound fragment is
rather resistant to proteolysis and is initially cleaved only at
Lys41, while cleavage of the Lys38-Glu39 peptide bond is
Spolaore et al.
expected to be hindered by the presence of an acidic residue
flanking the peptide bond to be hydrolyzed by trypsin (61).
Therefore, considering that Arg8 is at or near the N-terminus
of helix 9-34 of intact hGH, limited proteolysis of the folded
fragment occurs in agreement with the boundaries of the
expected helix. On the opposite, proteolysis of the fragment
alone is very fast and leads to the preferential hydrolysis at
the level of residues 8, 16, and 19, indicating that the helix
is not present.
The heme binding to the fragment likely involves, besides
a simple docking of heme to the hydrophobic face of the
amphiphilic helix of the fragment, also some specific ligands
that make the complex stronger and more specific. The UV-
visible absorption spectra of the complex show characteristics
of a hexacoordinated heme iron (see Results). Fragment
1-44 contains His18, His21, and Met14 (Figure 9), i.e.,
amino acid residues whose side chains are known to be
favored ligands in cytochromes (64). Both bis-His and Met/
His ligations occur in cytochromes, with His/His coordination
predominating both in the b-type and in c-type cytochromes.
The Met/His heme iron ligation occurs in the covalently
attached heme of horse cytochrome c (65), as well as in the
noncovalently bound heme of the four-helix bundle protein
cytchrome b562 (66) and in the fungal flavocytochrome
cellobiose dehydrogenase (67). Therefore, it can be proposed
that a bis-His or a Met/His complex is formed by fragment
1-44 using two His or one His and Met14, respectively, as
ligands. Unfortunately, the UV-visible absorption spectra
of oxidized and reduced heme bound to fragment 1-44
cannot be used to distinguish whether a bis-His or Met/His
complex is formed, since the two types of complexes have
similar maxima in their UV-visible absorption spectra (20,
68). The weak absorption band at 695 nm, often regarded
as diagnostic for the methionine ligation to the heme iron,
was not observed with the heme-fragment complex. How-
ever, the absence of this band does not necessarily indicate
that a methionine ligand is missing (see 69). Perhaps, even
the Tyr residues of the fragment 1-44 (Tyr28, Tyr35, and
Tyr42) can play a role in heme binding, since Tyr is weakly
nucleophilic and it acts as a heme ligand in Serratia
hemophore and in few other proteins (see ref 70 for
references). However, the very easy and fast proteolytic
cleavage of the C-terminal portion of the heme-fragment
complex seems to exclude at least Tyr42 as a heme ligand
(see Figure 8).
Resonance Raman (RR) measurements were conducted on
the heme-fragment complex, hoping to possibly determine
the amino acid side chains of fragment 1-44 which
coordinate heme. The characteristics of the RR spectra were
consistent with an Fe(III) in a hexacoordinate low spin state,
but the spectra did not enable us to distinguish between a
bis-His or a Met/His axial heme coordination (data not
shown). Perhaps, NMR studies could be conducted in order
to identify the heme coordinating residues, but the low
solubility of fragment 1-44 does not allow one to reach the
fragment concentration required for NMR measurements.
Therefore, additional studies are required in order to firmly
ascertain the ligands in the heme-fragment complex.
It is not known yet if heme binding to hGH fragment 1-44
does occur in vivo and thus if this binding has a physiological
significance. Heme in vivo is produced in the plasma upon
rupture of the blood cells, as a result of hemolysis, trauma,
Heme Binding by the hGH Fragment 1-44 Biochemistry, Vol. 44, No. 49, 2005 16087

FIGURE 9: Structural features of fragment 1-44. (Top) Amino acid sequence of fragment 1-44. (Bottom) Three-dimensional structure (A)
of fragment 1-44 of hGH derived from the X-ray structure of the protein (Brookhaven Protein Data Bank 3HHR) using the program
WebLab Viewer Pro 4.0 (Molecular Simulations Inc., San Diego, CA). The first helix of hGH (residues 9-34) is colored in red, and amino
acid residues relevant to the discussion of the results of this study are colored in blue (His) and green (Met). (B) The region 9-31 of helix
9-34 is shown looking through the helix axis. Hydrophobic and hydrophilic residues are colored in blue and in red, respectively, in order
to illustrate the amphiphilicity of the helix. (C) The same region 9-31 of the helix is also represented in a helix wheel drawing, and the
hydrophobic side of the helix is separated from the hydrophilic one by a dashed line.

and ischemia. It is currently accepted that free heme does
not exist in vivo, since it has toxic effects, but that it is
strongly bound to hemopexin and albumin. Hemopexin is
the strongest heme-binding protein in plasma (Kb 109 M-1)
(10), but its abundance is low (10-20 µM) and, therefore,
a protective effect from heme can be exerted by other heme-
binding proteins. It has been proposed that the most abundant
protein albumin can serve as a heme-carrier, when he-
mopexin is saturated (71). At present, we can speculate that,
if heme binding occurs, the hGH fragment adopts a folded
structure that makes it rather resistant to proteolytic degrada-
tion. The properties of fragment 1-44, produced in vitro by
proteolysis of the hormone (39), are expected to be very
similar to those of fragment 1-43, which is circulating in
vivo and shows hypoglycemic activity (33, 34, 38). If we
accept this, it is quite unlikely that the naturally occurring
fragment 1-43 circulates in vivo as a largely unfolded chain,
since it can be anticipated that this random polypeptide would
be degraded rapidly by proteases. Therefore, a stabilization
of this largely unfolded fragment by a mechanism of binding
and folding is necessary, as earlier proposed for intrinsically
unstructured or “natively unfolded” proteins (72).
Summing up, the interest of this study resides in the fact
that heme-binding is observed with a polypeptide species
that appears to be physiologically relevant. The hGH
fragment 1-44 can be an interesting model for further studies
of the molecular mechanisms of heme binding to polypep-
tides. Indeed, nowadays numerous investigators are involved
in studying a variety of model systems in an attempt to
provide insights into structure-function relationships of
heme-protein interactions and reproduce in relatively small
polypeptides the biological properties of the natural heme-
proteins that in vivo catalyze some vital bioenergetic
reactions (17, 20, 28, 73). Finally, we may add here that
perhaps the observation that a largely unfolded polypeptide
chain can bind heme and that binding results in the induction
16088 Biochemistry, Vol. 44, No. 49, 2005
and stabilization of peptide structure can have biomedical
implications, as follows. It has been found that partly folded
and “natively unfolded” proteins can be intermediates in the
protein aggregation processes that cause severe neurodegen-
erative diseases (Parkinson, Alzheimer) (74, 75). Heme
binding can occur with unstructured proteins, preventing
protein aggregation and ultimately the formation of amyloid
deposits (76-79). Therefore, the ability of heme to induce
and stabilize protein structure can be a mechanism involved
in the inhibition of protein aggregation.
ACKNOWLEDGMENT
We thank Dr. Patrizia Polverino de Laureto for insightful
comments on this paper and Marcello Zambonin for excellent
technical assistance. We acknowledge Prof. Angela Lombardi
and Prof. Giulietta Smulevich for important suggestions and
Dr. Barry D. Howes for performing resonance Raman
spectroscopy measurements.
REFERENCES
1. Dolphin, D., Ed. (1979) The Porphyrins, Vol. 7, Academic Press,
New York.
2. Padmanaban, G., Venkateswar, V., and Rangarajan, P. N. (1989)
Heme as a multifunctional regulator, Trends Biochem. Sci. 14,
492-496.
3. Feng, Y., Sligar, S. G., and Wand, A. J. (1994) Solution structure
of apocytochrome b562, Nat. Struct. Biol. 1, 30-35.
4. Falzone, C. J., Mayer, M. R., Whiteman, E. L., Moore, C. D.,
and Lecomte, J. T. (1996) Design challenges for hemoproteins:
The solution structure of apocytochrome b5, Biochemistry 35,
6519-6526.
5. Eliezer, D., and Wright, P. E. (1996) Is apomyoglobin a molten
globule? Structural characterization by NMR, J. Mol. Biol. 263,
531-538.
6. Robinson, C. R., Liu, Y., Thomson, J. A., Sturtevant, J. M., and
Sligar, S. G. (1997) Energetics of heme binding to native and
denatured states of cytochrome b562, Biochemistry 36, 16142-
16146.
7. Dumont, M. E., Corin, A. F., and Campbell, G. A. (1994)
Noncovalent binding of heme induces a compact apocytochrome
c structure, Biochemistry 33, 7368-7378.
8. Spolaore, B., Bermejo, R., Zambonin, M., and Fontana, A. (2001)
Protein interactions leading to conformational changes monitored
by limited proteolysis: Apo form and fragments of horse
cytochrome c, Biochemistry 40, 9460-9468.
9. Baltzer, L. (1998) Functionalization of designed folded polypep-
tides, Curr. Opin. Struct. Biol. 8, 466-470.
10. Hrkal, Z., Vodrazka, Z., and Kalousek, I. (1974) Transfer of heme
from ferrihemoglobin and ferrihemoglobin isolated chains to
hemopexin, Eur. J. Biochem. 43, 73-78.
11. Leclerc, E., Leclerc, L., Poyart, C., and Marden, M. C. (1993)
Interaction of heme with amphiphilic peptides: Use of hemin-
CN to probe the interaction of calmodulin with its target peptides,
Arch. Biochem. Biophys. 306, 158-162.
12. Marden, M. C., Dufour, E., Christova, P., Huang, Y., Leclerc, E.,
and Haertlé, T. (1994) Binding of heme-CO to bovine and porcine
β-lactoglobulin, Arch. Biochem. Biophys. 311, 258-262.
13. Hargrove, M. S., Barrick, D., and Olson, J. S. (1996) The
association rate constant for heme binding to globin is independent
of protein structure, Biochemistry 35, 11293-11299.
14. Zhang, L., and Guarente, L. (1995) Heme binds to a short sequence
that serves a regulatory function in diverse proteins, EMBO J.
14, 313-320.
15. Wardell, M., Wang, Z., Ho, J. X., Robert, J., Ruker, F., Ruble, J.,
and Carter, D. C. (2002) The atomic structure of human Meth-
emalbumin at 1.9 Å, Biochem. Biophys. Res. Commun. 291, 813-
819.
16. Poulos, T. (1995) The role of proximal ligand in heme enzymes,
J. Bioinorg. Chem. 1, 356-359.
17. Reedy, C. J., and Gibney, B. R. (2004) Heme protein assemblies,
Chem. ReV. 104, 617-649.
Spolaore et al.
18. Nastri, F., Lombardi, A., D’Andrea, L. D., Sanseverino, M.,
Maglio, O., and Pavone, V. (1998) Miniaturized hemoproteins,
Biopolymers 47, 5-22.
19. Rau, H. K., and Haehnel, W. (1998) Design, synthesis and
properties of a novel cytochrome b model, J. Am. Chem. Soc.
120, 468-476.
20. Lombardi, A., Nastri, F., and Pavone, V. (2001) Peptide-based
heme-protein models, Chem. ReV. 101, 3165-3189.
21. Gibney, B. R., and Dutton, P. L. (2000) De noVo design and
synthesis of heme proteins, in AdVances in Inorganic Chemistry
(Mauk, A. G., and Sykes, A. G., Eds.) Vol. 51, pp 409-455,
Academic Press, London.
22. Hecht, M. H., Richardson, J. S., Richardson, D. C., and Ogden,
R. C. (1990) De noVo design, expression and characterization of
Felix: A four-helix bundle protein of native-like sequence, Science
249, 884-891.
23. Robertson, D. E., Farld, R. S., Moser, C. C., Urbauer, J. L.,
Mulholland, S. E., Pidikiti, R., Lear, J. D., Wand, A. J., De Grado,
W. F. and Dutton, P. L. (1994) Design and synthesis of multi-
heme proteins, Nature 368, 425-432.
24. Choma, C. T., Lear, J. D., Nelson, M. J., Dutton, P. L., Robertson,
D. E., and De Grado, W. F. (1994) Design of a heme-binding
four-helix bundle, J. Am. Chem. Soc. 116, 856-865.
25. Rojas, N. R. L., Kamtekar, S., Simone, C. T., Mc Lean, J. E.,
Vogel, K. H., Spiro, T. G., Farid, R. S., and Hecht, M. H. (1997)
De noVo heme proteins from designed combinatorial libraries,
Protein Sci. 6, 2512-2524.
26. Gibney, B. R., Rabanal, F., Reddy, K. S., and Dutton, P. L. (1998)
Effect of four-helix bundle topology on heme binding and redox
properties, Biochemistry 37, 4635-4643.
27. De Grado, W. F., Summa, C. M., Pavone, V., Nastri, F., and
Lombardi, A. (1999) De noVo design and structural characteriza-
tion of proteins and metalloproteins, Annu. ReV. Biochem. 68,
779-819.
28. Ghirlanda, G., Osyczka, A., Liu, W., Antolovich, M., Smith, K.
M., Dutton, P. L., Wand, A. J., and De Grado, W. F. (2004) De
noVo design of a D2-symmetrical protein that reproduces the
diheme four-helix bundle in cytochrome cb1, J. Am. Chem. Soc.
126, 8141-8147.
29. de Vos, A. M., Ultsch, M., and Kossiakoff, A. A. (1992) Human
growth hormone and extracellular domain of its receptor: Crystal
structure of the complex, Science 255, 306-312.
30. Li, C. H. (1982) Human growth hormone: 1974-1981, Mol. Cell.
Biochem. 46, 31-41.
31. Silva, C. M., Isgaard, J., and Thorner, M. O. (1998) Cytokines in
endocrine function, AdV. Protein Chem. 52, 199-221.
32. Okada, S., and Kopchick, J. J. (2001) Biological effects of growth
hormone and its antagonist, Trends Mol. Med. 7, 126-132.
33. Lewis, U. J., Sinha, Y. N., and Lewis, G. P. (2000) Structure and
properties of members of the hGH family: A review, Endocr. J.
47, Suppl., S1-S8.
34. Singh, R. N. P., Seavey, B. K., Lewis, L. J., and Lewis, U. J.
(1983) Human growth hormone peptide 1-43: Isolation from
pituitary glands, J. Protein Chem. 2, 425-436.
35. Frigeri, L. G., Teguh, C., Ling, N., Wolff, G. L., and Lewis, U. J.
(1988) Increased sensitivity of adipose tissue to insulin after in
vivo treatment of yellow Avy/A obese mice with amino-terminal
peptides of human growth hormone, Endocrinology 122, 2940-
2945.
36. Lewis, U. J., Lewis, L. J., Salem, M. A., Staten, N. R., Galosy, S.
S., and Krivi, G. G. (1991) A recombinant-DNA-derived modi-
fication of human growth hormone (hGH44-191) with enhanced
diabetogenic activity, Mol. Cell. Endocrinol. 78, 45-54.
37. Sinha, Y. N., and Jacobsen, B. P. (1994) Human growth hormone
(hGH)-(44-191), a reportedly diabetogenic fragment of hGH,
circulates in human blood: Measurement by radioimmunoassay,
J. Clin. Endocrinol. Metab. 78, 1411-1418.
38. Lopez-Guajardo, C. C., Armstrong, L. S., Jordan, L., Staten, N.
R., Krivi, G. G., Martinez, A. O., and Haro, L. S. (1998)
Generation, characterization and utilization of anti-human growth
hormone 1-43 (hGH1-43), monoclonal antibodies in an ELISA,
J. Immunol. Methods 215, 179-185.
39. Spolaore, B., Polverino de Laureto, P., Zambonin, M., and Fontana,
A. (2004) Limited proteolysis of human growth hormone at low
pH: Isolation, characterization and complementation of the two
biologically relevant fragments 1-44 and 45-191, Biochemistry
43, 6576-6586.
40. Franchi, E, Maisano, F., Testori, S. A., Galli, G., Toma, S., Parente,
L., de Ferra, F., and Grandi, G. (1991) A new human growth
Heme Binding by the hGH Fragment 1-44
hormone production process using a recombinant Bacillus subtilis
strain, J. Biotechnol. 18, 41-54.
41. Gill, S. C., and von Hippel, P. H. (1989) Calculation of protein
extinction coefficients from amino acid sequence data, Anal.
Biochem. 182, 319-326.
42. Dawson, R. M. C., Elliott, D. C., Elliott, W. H., and Jones, K. M.
(1975) Data for Biochemical Research, pp 230-231, Oxford
University Press, Oxford, U.K.
43. Tsutsui, K., and Mueller, G. C. (1982) A protein with multiple
heme-binding sites from rabbit serum, J. Biol. Chem. 257, 3925-
3931.
44. Kuwabara, T., Nakamura, A., Ueno, A., and Toda, F. (1994)
Inclusion complexes and guest-induced color changes of pH-
indicator-modified β-cyclodextrins, J. Phys. Chem. 98, 6297-
6303.
45. Sakamoto, S., Sakurai, S., Ueno, A., and Mihara, H. (1997) Heme
binding and catalytic activity of two-R-helix peptide annealed by
trifluoroethanol, Chem. Commun. 13, 1221-1222.
46. Scholtz, J. M., Qian, H., York, E. J., Stewart, J. M., and Baldwin,
R. L. (1991) Parameters of helix-coil transition theory for alanine-
based peptides of varying chain lengths in water, Biopolymers
31, 1463-1470.
47. Santoni, E., Scatragli, S., Sinibaldi, F., Fiorucci, L., Santucci, R.,
and Smulevich, G. (2004) A model for the misfolded bis-His
intermediate of cytochrome c: The 1-56 N-fragment, J. Inorg.
Biochem. 98, 1067-1077.
48. Babcock, G. T., Widger, W. R., Cramer, W. A., Oertling, W. A.,
and Metz, J. G. (1985) Axial ligands of chloroplast cytochrome
b-559: Identification and requirement for a heme-crosslinked
polypeptide structure, Biochemistry 24, 3638-3645.
49. Kelly, S. M., Jess, T. J., and Price, N. C. (2005) How to study
proteins by circular dichroism, Biochim. Biophys. Acta 1751, 119-
139.
50. Benson, D. R., Hart, B. R., Zhu, X., and Doughty, M. B. (1995)
Design, synthesis, and circular dichroism investigation of peptide-
sandwiched mesoheme, J. Am. Chem. Soc. 117, 8502-8510.
51. Nicola, N. A., Minasian, E., Appleby, C. A., and Leach, S. J.
(1975) Circular dichroism studies of myoglobin and leghemoglo-
bin, Biochemistry 14, 5141-5149.
52. Myer, Y. P. (1978) Circular dichroism spectroscopy of emopro-
teins, Methods Enzymol. 54, 249-284.
53. Dockal, M., Carter, D. C., and Ruker, F. (1999) The three
recombinant domains of human serum albumin: Structural
characterization and ligand binding properties, J. Biol. Chem. 274,
29303-29310.
54. Ihara, M., Takahashi, S., Ishimori, K., and Morishima, I. (2000)
Functions of fluctuation in the heme-binding loops of cytochrome
b5 revealed in the process of heme incorporation, Biochemistry
39, 5961-5970.
55. Hernandez, H., and Robinson, C. V. (2001) Dynamic protein
complexes: Insights from mass spectrometry, J. Biol. Chem. 276,
46685-46688.
56. Sobott, F., and Robinson, C. V. (2002) Protein complexes gain
momentum, Curr. Opin. Struct. Biol. 12, 729-734.
57. He, F., Hendrickson, C. L., and Marshall, A. G. (2000) Unequivo-
cal determination of metal atom oxidation state in naked heme
proteins: Fe(III)myoglobin, Fe(III)cytochrome c, Fe(III)cyto-
chrome b5, and Fe(III)cytochrome b5 L47R, J. Am. Soc. Mass
Spectrom. 11, 120-126.
58. Fontana, A., Fassina, G., Vita, C., Dalzoppo, D., Zamai, M., and
Zambonin, M. (1986) Correlation between sites of limited
proteolysis and segmental mobility in thermolysin, Biochemistry
25, 1847-1851.
59. Fontana, A., Zambonin, M., Polverino de Laureto, P., De Filippis,
V., Clementi, A., and Scaramella, E. (1997) Probing the confor-
mational state of apomyoglobin by limited proteolysis, J. Mol.
Biol. 266, 223-230.
Biochemistry, Vol. 44, No. 49, 2005 16089
60. Fontana, A., Polverino de Laureto, P., Spolaore, B., Frare, E.,
Picotti, P., and Zambonin, M. (2004) Probing protein structure
by limited proteolysis, Acta Biochim. Pol. 51, 299-321.
61. Ambler, R. P., and Brown, L. H. (1967) The amino acid sequence
of Pseudomonas fluorescent azurin, Biochem. J. 104, 784-825.
62. Huffman, D. L., and Suslik, K. S. (2000) Hydrophobic interactions
in metalloporphyrin-peptide complexes, Inorg. Chem. 39, 5418-
5419.
63. Huffman, D. L., Rosenblatt, M. M., and Suslick, K. S. (1998)
Synthetic heme-peptide complexes, J. Am. Chem. Soc. 120, 6183-
6184.
64. Moore, G. R., and Pettigrew, G. W. (1990) Cytochromes c:
EVolutionary, Structural and Physicochemical Aspects, Springer-
Verlag, New York.
65. Bushnell, G. W., Louie, G. V., and Brayer, G. D. (1990) High-
resolution three-dimensional structure of horse heart cytochrome
c, J. Mol. Biol. 214, 585-595.
66. Hay, S., and Wydrzynski, T. (2005) Conversion of the Escherichia
coli cytochrome b562 to an archetype cytochrome b: A mutant
with bis-histidine ligation of heme iron, Biochemistry 44, 431-
439.
67. Rotsaert, F. A., Hallberg, B. M., de Vries, S., Moenne-Loccoz,
P., Divne, C., Renganathan, V., and Gold, M. H. (2003) Biophysi-
cal and structural analysis of a novel heme B iron ligation in the
flavocytochrome cellobiose dehydrogenase, J. Biol. Chem. 278,
33224-33231.
68. Ishida, M., Dohmae, N., Shiro, Y., Oku, T., Iizuka, T., and Isogai,
Y. (2004) Design and synthesis of de noVo cytochromes c,
Biochemistry 43, 9823-9833.
69. Rosell, F. I., and Mauk, A. G. (2002) Spectroscopic properties of
a mitochondrial cytochrome c with a single thioether bond to the
heme prosthetic group, Biochemistry 41, 7811-7818.
70. Deniau, C., Gilli, R., Izadi-Pruneyre, N., Letoffe, S., Delepierre,
M., Wandersman, C., Briand, C., and Lecroisey, A. (2003)
Thermodynamics of heme binding to the HasA(SM) hemophore:
Effect of mutations at three key residues for heme uptake,
Biochemistry 42, 10627-10633.
71. Peters, T., Jr. (1996) All about Albumins: Biochemistry, Genetics
and Medical Applications, Academic Press, San Diego, CA.
72. Wright, P. E., and Dyson, H. J. (1999) Intrinsically unstructured
proteins: Re-assessing the protein structure-function paradigm,
J. Mol. Biol. 293, 321-331.
73. Rosenblatt, M. M., Wang, J., and Suslick, K. S. (2003) De noVo
designed cyclic-peptide heme complexes, Proc. Natl. Acad. Sci.
U.S.A. 100, 13140-13145.
74. Dobson, C. M. (2003) Protein folding and disease: A view from
the first Horizon Symposium, Nat. ReV. Drug DiscoV. 2, 154-
160.
75. Uversky, V. N., and Fink, A. L. (2004) Conformational constraints
for the amyloid fibrillation: The importance of being unfolded,
Biochim. Biophys. Acta 1698, 131-153.
76. Howlett, D., Cutler, P., Heales, S., and Camilleri, P. (1997) Hemin
and related porphyrins inhibit beta-amyloid aggregation, FEBS
Lett. 417, 249-251.
77. Caughey, W. S., Raymond, L. D., Horiuchi, M., and Caughey, B.
(1998) Inhibition of protease-resistant prion protein formation by
porphyrins and phthalocyanines, Proc. Natl. Acad. Sci. U.S.A. 95,
12117-12122.
78. Priola, S. A., Raines, A., and Caughey, W. S. (2000) Porphyrin
and phthalocyanine antiscrapie compounds, Science 287, 1503-
1506.
79. Pato, C., Celier, C., Rezael, H., Grosclaude, J., and Marden, M.
C. (2004) Heme as an optical probe of conformational transition
of ovine recPrP, Protein Sci. 13, 1100-1107.
BI051374D